Data helps businesses improve their strategy

by Laura Scavino

Many organisations started looking at customer databases to take advantage of the great amount of information available to them. The correct use of data analytics tools helps businesses  better targeting existing customers and also qualifying potential new ones. Data analysis can not only lower marketing costs but it also maintain and increase revenues while reaching new customers and retaining the existing ones.

Database marketing is usually usually used for selling products or services to new customers, selling new products or services to existing customers and to monitor relationships with existing customers. For these reasons software tools and models can be used to summarise large amounts of data in order to support decision making and strategy implementations. It can be way more effective to use modelling tools rather than direct marketing because the tools help businesses targeting the most relevant customers.

Predictive modelling tools is different from marketing segmentation analysis because rather than segmenting customers by age, gender, purchase history; predictive modelling look instead at insight-derived needs and analysis of customer responses.

Among the predictive modelling tools the linear and logistic regression are two of the most important tools of traditional statistical inference. For the linear regression, the linear predictor makes the prediction, while in logistic regression the linear predictor is edited in a way that the predicted probability for each category of variables is between zero and one and the sum of probabilities across categories is one.

Independent Variables = Predictor Variables = Inputs to predict an outcome

Dependent Variable = Target Variable = Outcome to predict with the inputs

Data investigation before the Predictive Analysis example in Alteryx 

-Association Analysis tool look at the bivariate association

 

 

-Histogram tool

 

 

-Distribution analysis tool shows which distribution best represent the data

 

-field summary provides a descriptive analysis and also an interactive report for the selected columns

Once the data has been investigated with the data investigation tools, it is easier to do predictive analysis in Alteryx. The data analysed looked at the energy price and consumption in all the US states. For the purpose of the analysis the sample chosen is 80-20 (estimation, validation), created with the sample tool in Alteryx.

Predictive tools in Alteryx

-Linear regression is a tool of traditional statistical inference used to constructs a linear function to create a model that predicts a target variable based on one or more predictor variables

 

-Boosted model creates a generalised boosted regression model based on gradient boosting methods. It determines which subset of fields best predicts a target field, it capture non-linear relationships and interactions between fields, it addresses problems

 

-Score tool determines the quality of a model’s predictions

 

To summarise, predictive modelling are vital analytical tools for businesses and they have great impact on strategic decisions.

What did I learn in the last 16 weeks at the Data School?

5 mins read

Sun 03 Jun 2018

Have a look at my top 5 Tableau Tips

5 mins read

Fri 01 Jun 2018

Create a grid map in Tableau

2 mins read

Tue 29 May 2018

SQL Fundamentals and examples on parallels with Tableau prep

4 mins read

Tue 22 May 2018